Tagraxofusp, azacitidine, and venetoclax (TAG-AZA-VEN) triplet therapy shows efficacy, tolerability, and transplant potential in patients with blastic plasmacytoid dendritic cell neoplasm (BPDCN): Results of a phase 2 trial

Abstract Background: BPDCN is an aggressive, orphan hematologic malignancy characterized by multi-organ involvement including skin, blood, bone marrow, lymph nodes, and central nervous system (CNS). A defining feature is high surface expression of CD123, the interleukin 3 receptor alpha chain. The only approved drug for BPDCN is tagraxofusp (TAG), comprised of recombinant IL3 coupled to a diphtheria toxin payload. TAG binds to CD123, resulting in targeted internalization and cytotoxicity. We previously evaluated safety of the triplet combination of TAG with AZA and VEN in a phase 1b study for AML (Lane et al, Blood Adv 2024). We also previously found that BPDCN is highly dependent on BCL2 and sensitive to VEN (Montero et al, Ca Disc 2017), and that resistance to TAG can be overcome by treatment with AZA (Togami et al, JCI 2019). Here, we performed a phase 2 study of TAG-AZA-VEN in patients with BPDCN (NCT03113643). Methods: Patients had previously untreated (1L) or relapsed/refractory (R/R) BPDCN. Eligibility was based on TAG BPDCN and AZA/VEN FDA labels, including normal ejection fraction and albumin >= 3.2 g/dL to mitigate risk of capillary leak syndrome (CLS). Treatment was 28-day cycles of AZA 75 mg/m2 days 1-7; VEN 400 mg days 1-21 (ramp up 100 mg d1, 200 mg d2, 400 mg d3-21 in cycle 1), and TAG 12 mcg/kg days 4, 5, 6. Response criteria included assessment of bone marrow, skin (mSWAT), and extramedullary sites (CT or PET/CT) as in the TAG registration trial (Pemmaraju et al, NEJM 2019). CNS evaluation at enrollment was mandatory; patients with asymptomatic CNS involvement were enrolled and received intrathecal chemotherapy. Results: 27 patients were enrolled, n=16 1L and n=11 R/R. Median age was 70 (range 21-81). As expected for BPDCN, 93% were male. 10 of 27 (37%) had “skin-only” disease; others had additional sites of organ involvement. 6 of 27 (22%) had asymptomatic CNS disease at study entry. In the R/R cohort (n=11), 7 had prior TAG and 4 prior transplant (3 allo, 1 auto). The median number of cycles received was 2.5 (range 1-14) in the 1L cohort, and 2 (range 1-5) in the R/R cohort. Median follow-up in the 1L cohort was 20.4 months (95% CI 6.4-26.8). In the R/R cohort, all patients died in follow-up except one who remains on study at 39 months. Treatment-related adverse events (TRAEs) were similar to those in the TAG-AZA-VEN AML study. Grade 3 or higher TRAEs: neutropenia (44%), thrombocytopenia (44%), anemia (11%), hypoxia (11%), atrial fibrillation (4%), febrile neutropenia (4%), sinus tachycardia (4%), hyperglycemia (4%), hypophosphatemia (4%). CLS occurred in 4 of 27 patients (15%); three grade 2, one grade 3, similar to rates with single agent TAG. All CLS events were in cycle 1 and were manageable. One 1L patient died during cycle 1 with multi-organ system failure. 30-day, 60-day mortality were 6%, 6% (1L) and 0%, 9% (R/R). In the 1L cohort, 14 of 16 patients (88%) achieved composite CR (CR, n=8; CRi, n=6), with one PR and one not evaluable (early death). In the R/R cohort, 7 of 11 patients (64%) achieved composite CR (CR, n=2; CRi, n=3; CRc, n=2). In the 1L cohort, median duration of response (DOR) was not reached. In the R/R cohort, median DOR was 7.2 months (95% CI 5.5-36.4). In the 1L cohort, 10 of 16 (63%) went to alloSCT directly following protocol therapy, including 10 of the 11 patients age <= 75. In the R/R cohort, 6 of 11 (55%) went to alloSCT. Median overall survival (OS) was not reached in the 1L cohort, with both 1- and 2-year OS of 60% (95% CI 32-88). Median progression-free survival (PFS) in the 1L cohort was not reached, with 1- and 2-year PFS of 58% (95% CI 32-84). In the R/R cohort, median OS was 8.4 months (95% CI 4.8-21.7), with 1-year OS of 36% (95% CI 8-65) and 2-year OS of 18% (95% CI 0-41). Median PFS in the R/R cohort was 6.3 months (95% CI 2.2-11). Conclusions: Triplet therapy with TAG-AZA-VEN is effective and feasible in patients with previously untreated or R/R BPDCN, including in an older patient population. The safety profile is consistent with prior studies of TAG and AZA/VEN, and triplet therapy does not add new AEs not seen with TAG or AZA/VEN. A high proportion of patients (1L and R/R) proceeded to transplant in remission. Response rates, bridge to transplant rates, and overall survival may improve upon that of single agent tagraxofusp, in both 1L and R/R patients (Pemmaraju et al, JCO 2022). These data support TAG-AZA-VEN as an effective treatment option for patients with BPDCN.